As correct application of published evaluation criteria will determine with whom an authority will contract for goods, services and works, this area merits extensive investment of time in the competition design phase. It is also the area most commonly scrutinised in procurement challenges, and scrupulous record-keeping will greatly improve the prospects of successfully defending a claim.
This is particularly important for above threshold contracts which engage TFEU principles. Such evaluation criteria must, for example :
In terms of disclosure of criteria, sub-criteria and weightings, a balance needs to be struck to enable effective competition: too much prescription as to how a question should be answered may stifle innovation between bids, eliminating any meaningful differentiation in response. On the other hand lack of prescription may confer unrestricted freedom of choice on the authority, enhances the risk of unequal treatment and of unsuitable proposals being put forward.
The guiding principles include the following:
An example of how the Court will apply a predictability test can be found in the case of Mears Ltd v Leeds City Council (see paragraph 122 here). In this case the Court held that the authority ought to have disclosed the fact that questions under each headline criteria were in fact weighted differently. The Court also had to consider whether a model answer template used by the evaluators to assess the strength of a response ought to have been disclosed. Whilst the Council had intended the model answers to be non-prescriptive guidance as an aid to consistent scoring and not part of the formal evaluation methodology, the Court found that in fact it was applied as part of the evaluation. The Court went through each element of the model answer systematically and asked whether it ought to have been foreseeable to Mears and whether it could have affected tenders. The Court found some instances of criteria which ought to have been disclosed (for example a question dealing with staff stipulated in the model answer that lists of staff should be provided). Examples which did not amount to undisclosed criteria included the requirement to address health and safety, and the requirement to address grants within energy efficiency. The Court held that aspects such as these ought to have been reasonably foreseeable to a contractor such as Mears.
It is not uncommon for the outcome of procurement evaluations to result in a few points (or even fractions) distinguishing the bids (particularly in developed markets in which offerings may be similar) so that the effect of undisclosed criteria could be decisive on the outcome.
4 months ago South West Council published an OJEU Notice for a new 3 year independent living mixed equipment and services contract, with the contract value for the supplies element exceeding the service element. The Council is using the restricted procedure. Following the PQQ process the Council invited tenders from 5 bidders. A 7 person panel assessed the written tenders and arrived at the following scores:
A Ltd 73%
B Ltd 73%
C Ltd 72%
D Ltd 63%
E Ltd 61%
The ITT stated that the Council reserved the right to invite bidders to an interview for the purposes of clarification of tenders, but that interviews would not be separately scored. The portfolio holder in charge of social services now wants to interview the top 3 placed bidders to see (a) what the teams delivering the service are like and (b) to understand the extent to which each bidder is employing "local labour" (to use the portfolio holder's words).
What legal considerations arise?
Can an interview be held at all? Unless an interview follows formulaic questions disclosed in advance, in practice it may be difficult to avoid the risk of introducing undisclosed factors and/or unequal treatment within an interview setting. For those procedures (including the restricted procedure) which do not permit any form of negotiation, interviews are best avoided.
It is however arguable that an interview affords an opportunity for a contracting authority to clarify aspects of a tender. In Case C 599/10 - SAG ELV Slovenkso a.s., and others v Úrad pre verejné obstarávanie, the ECJ set out the principles to be applied when dealing with an imprecise tender. However this may not amount to the submission of a new or improved tender. However, authorities may prefer to conduct their clarification exercises in writing in order to have a clear audit trail of particular items which merit clarifications and what responses are being accepted as forming part of the tender.
If an interview is used, careful consideration therefore needs to be given to what questions are to be asked, and whether this amounts to clarification of criteria already disclosed or whether it introduces new criteria. Careful consideration would also need to be given if not inviting all tenderers to an interview as this may exacerbate the risk of unequal treatment.
For example, if the purpose of the interview is to ascertain the extent to which the proposed staff team will communicate well with disabled service users, is this criterion part of the disclosed evaluation matrix, or at least it is reasonably foreseeable that it is?
Assessing the staff team: One concern may be that an interview strays into what previous experience a bidder has. The distinction between selection criteria and evaluation criteria as set out in Lianakis must be observed (Lianakis v Dimos Alexandroupolis (Case C-532/06)). However, recent direction of travel by the ECJ indicates that past experience can be lawful evaluation criteria where it is properly used to assess ability to deliver this contract .
This direction of travel is now encapsulated within the Public Contracts Regulations 2015: Criteria which relate to the organisation, qualification and experience of staff are expressly permitted by Regulation 67 where the quality of staff can have a significant impact on the level of performance of the contract.
"Local Labour": Regulation 67 also permits environmental and social criteria to be taken into account, provided it is linked to the subject matter of the contract. In the Beentjes decision  the ECJ held that a social criterion which required the employment of long-term unemployed from the region could be used as a tool for implementing social policies, providing that the principles of non-discrimination and transparency were observed . Care must be taken to avoid discriminatory requirements: for example, a requirement to use Danish labour and Danish materials was held to be incompatible with TFEU principles in the Storebaelt case . Similarly, stipulating that a provider must have an established office in the area at the time the bid is submitted was held to be unlawful in Contse SA .
Authorities may wish to lay down special conditions relating to the performance of the proposed contract and evaluate tenderers' ability to comply with those conditions. Under Article/Regulation 70 this is permissible provided that the condition is linked to the subject matter of the contract and indicated in the procurement documents. The recent ECJ ruling in RegioPost  held that a mandatory requirement that all tenderers certify they would pay the local minimum wage was non-discriminatory, had been transparently referred to in the procurement documents, and met the requirements of Article 26 of the 2004 Directive (the precursor to Article 70).
Service Users: Within the context of the Public Contracts Regulations 2006, the High Court has confirmed that there is no express prohibition on service users or their carers forming part of an evaluation panel for a tender . However, this does not mean that service users can influence scoring based on undisclosed criteria. Advance consideration of how service users' views can be taken into account will enable adequate disclosure of this criterion and help ensure equal treatment.
Indeed, Regulation 76 of the Public Contracts Regulations 2015 (as amended) recognises that the involvement and empowerment of users could be an award criteria for "light regime" procurements.
A number of recent cases have demonstrated the importance of an adequate audit trail, and how the absence of notes and ability to explain the evaluation process can count against a contracting authority. In this section we consider a number of practical questions which arise in creating a proper audit trail.
Should evaluators mark in a silo, or mark together? Occasionally we hear the view that evaluators are told to mark in a room all together, to eliminate the risk of a rogue evaluator making adverse comments or notes. The preferred view is that evaluators should be properly trained so as to mitigate this risk, and that marking together does not afford each evaluator the independence of thought that case law requires .
Evaluators should be trained adequately in the general principles, the way of evaluation methodology ought to be applied and the scoring system. It should be explained that they must document their reasons carefully against the evaluation criteria, referring to the content of the bid and not just echo the language used in the scoring matrix. Also, in the event that an evaluator considers a tenderer to have provided irrelevant information, the evaluator ought to say so in the reasons given for the score. They should be aware that if there is a challenge, their comments may be scrutinised and the importance of the role they are carrying out.
Authorities should also ensure that no conflicts of interest arise (see Regulation 24 for what can be termed a conflict of interest) and that evaluators are aware of the requirements of Regulation 24 regarding conflicts of interest. In Counted4 Community Interest Company v Sunderland City Council the Court held that there was serious issue to be tried in relation to the involvement of a contract manager on the evaluation panel, whose relationship which the incumbent claimant had soured beyond normal managerial issues, and refused to lift the suspension.
Should the authority keep all notes, including the initial notes of individual evaluators? Some clients express a view that only final notes should be kept (such as the notes from final moderation meetings). This is no doubt born out of a desire to ensure that only lawful and considered reasons form part of the audit trail. The risk of an incomplete audit trail is that it can be very difficult to disprove an allegation of (for example) bias, manifest error or proper moderation if it is impossible to demonstrate how each evaluator reached its initial view. Further, in the Resource NI case the Court held that:
"…meetings of contract procurement panels are something considerably greater than merely formal events. They are solemn exercises of critical importance to economic operators and the public and must be designed, constructed and transacted in such a manner to ensure that full effect is given to the overarching procurement rules and principles. Where, in any given case, a disappointed bidder's legal challenge focuses on the activities and deliberations of an evaluation panel, the evidence bearing thereon will, inevitably, be carefully and objectively scrutinised by the court."
Furthermore, Regulation 84 requires contracting authorities to ensure that they keep sufficient documentation to justify decisions taken in all stages of the procurement procedure, including the evaluation stage. In practical terms, on-line systems often require each evaluator to log into the system individually and enter a score with supporting reasons for each question. The evaluators cannot see each other's scores/reasons at this stage.
How should marks be moderated? The Regulations do not contain an express requirement to moderate, but moderation helps ensure that the final decision and reasons for the award of a contract follow TFEU principles and the Regulations.
There is no prescribed method for moderation, and different practices include moderation by consensus (either via sequential discussion with individual evaluators or simultaneous discussion collectively in one room), and moderation by review alone (rather than through discussion). This latter method has been adopted in relation to very large procurement exercises (such as the two-tier tender run by the Legal Aid Agency which was subject of a judicial review). One problem with adopting this approach is that it risks reducing the independence of thought which was delegated to the individual evaluators in the first place. Also, the evaluation should result in a unified consensus view to which all panel members subscribe (see again, Resource NI) and a group moderation facilitates this.
Variation in practice also occurs in relation to how the authority arrives at the final score. Some authorities adopt the approach of averaging scores. Some authorities adopt the approach of removing outliers and arriving at a median score. Other authorities adopt the approach of arriving at a consensus score. The benefit of the latter approach is that it enables views to be checked and discussed, and a consensus organisational score to be reached. The risk of removing outliers is again that it removes independence of thought. Whilst averaging is common is does not benefit from the organisational view as it is necessarily a mathematical construct. It could also lead to conflicting debriefing statements since there may not be a consistent articulation of the characteristics and advantages of the winning tender which aligns with the final score awarded.
Occasionally even the most robust of moderation procedures will not be enough: For example in Woods Building Service v Milton Keynes Council  EWHC 2011 (TCC) the Court held that the review exercise conducted by the Council did not remedy the defects in the initial evaluation. The Council ought to have started the evaluation again from scratch (see paragraph 41).
1. All references are to the Public Contracts Regulations 2015 unless otherwise stated. See Regulations 18, 49, 56, 67. See also Directive 2012/24/EU Articles 56 and 67
2. Case C-19/00 - SIAC Construction Ltd v County Council of the County of Mayo
3. Case C-601/13 Ambisig v Nersent; Case T 447/10, Evropaïki Dynamiki – Proigmena Systimata Tilepikoinonion Pliroforikis kai Tilematikis AE v Court of Justice of European Union
4. Case 31/87, Gebroeders Beentjes v The Netherlands
5. See also Nord-Pas-de-Calais case C-225/98, European Communities v French Republic
6. C-234/89 Storebaelt
8. AJ and another v Calderdale Primary Care Trust  EWHC 3552 (Admin)
9. Resource (NI) v Northern Ireland Courts & Tribunals Service  NIQB 121
10. Resource (NI) (ibid) at paragraph 39(b)
11.  EWHC 3898 (TCC)
At what stage in a tender process must a tenderer provide proof that products which it proposes to deliver are equivalent to those...